mbetaTopN

Syntax

mbetaTopN(X, Y, S, window, top, [ascending=true], [tiesMethod=’oldest’])

Please see Moving TopN Functions (mTopN functions) for the parameters and windowing logic.

Details

After stably sorting S in the specified ascending order, the function obtains the first top pairs of elements in X and Y in the sliding window and calculates the coefficient estimate ordinary-least-squares regressions of Y on X.

Examples

$ x = NULL 3 8 4 0
$ y = 2 3 1 7 3
$ s = 5 NULL 8 9 4
$ mbetaTopN(x, y, s, 3, 2)
[ , , , -0.66, -4]

$ s2=2021.01.01 2021.02.03 2021.01.23 2021.04.06 2021.12.29
$ mbetaTopN(x, y, s2, 3, 2)
[ , , , -2.5, -0.6667]

$ x1 = matrix(x, 4 3 6 2 3)
$ y1=matrix(3 7 9 3 2, y)
$ s1=matrix(2 3 1 7 3, s)

$ mbetaTopN(x, y1, s1, 3, 2)

col1	col2
2.5	-2.5
1.1429	-4

$ mbetaTopN(x1, y1, s, 3, 2)

col1	col2
	-1
	-1
2.5	-1.5
1.1429	-1.5

$ mbetaTopN(x1, y1, s, 3, 2)

col1	col2
	-1
	-1
2.5	-1.5
1.1429	-1.5

$ n = 3000
$ ids = 1..3000
$ dates = take(2021.01.01..2021.10.01,n)
$ prices = rand(1000,n)
$ vals = rand(1000,n)
$ t = table(ids as id,dates as date,prices as price,vals as val)
$ dbName = "dfs://test_mbetaTopN_2"
$ if(existsDatabase(dbName))dropDB(dbName)
$ db = database(dbName,VALUE,1..5000)
$ pt = db.createPartitionedTable(t,"pt",`id).append!(t)
$ select mbetaTopN(price, val, id, 10, 5, true) from pt where date>2021.05.01

Related function: mbeta